Google envisions a future where almost every interaction possible over the Internet is personalized.
According to Google’s VP of product search, Robby Stein, most people use AI products not for general information but more to come up with advice and recommendations.
“They want to know where they’re going to have dinner, they want to know where they’re going to go with their family,” Stein told the Endless podcast last week. “So we think there’s a good opportunity for our AI to know itself better and help differently because of that knowledge.”
In order for AI to get a better understanding of you, Stein says that Google’s AI models will use connected services like Gmail to be able to paint a more detailed picture of your likes and dislikes over time. As a result, for example, AI can serve you a targeted product that you might like.
AI is a trending topic for google searches. Earlier this year, a judge ruled against Google in an antitrust investigation of its search business, saying that for the first time in ten years, “it’s the first time” a product has come out that will present a meaningful challenge to Google’s market dominance. ” It seems that Google, may see where the judge is coming from as the Tech Giant breaks through its AI efforts, unveiling its latest Gemini model to great success for the entire Google Ecosystem.
Gemini is already integrated with Google Workspace applications such as Gmail (where personal notes), calendar (which knows what you are doing, when, which can contain your work documents or your photos). It’s also available on Google Maps, YouTube, WhatsApp, Spotify, you name it.
The company also launched a new Chrome browser in September with gemini integration throughout, including agentic ai that can navigate the web and complete tasks on your behalf.
Stein’s vision of an Internet that knows more about you is when your interactions with Google will not be limited to the specific contexts you use the search engine, but rather ongoing conversations. The example he gives, he says, is you want a bed for your apartment. You will gradually learn your AI information about what kind of bed you would like, whenever it pops into your head. AI will remember these scattered pieces of information, and much more that it has gathered from you through past conversations and your activity in connected apps. And then one day, the perfect bed you want goes on sale, and Boom, the AI uses it for you, perhaps with a careful alending.
“I think that’s a lot more how I think about the future of search than any one aspect or aspect of the form,” Stein said.
Of course, some parts of the Google Search experience would remain arbitrary, even in this view, such as when you go to the Internet to look up simple, factual information, such as the height of the Empire State Building. But that’s probably the majority, according to a Google executive.
“It’s almost weird that it doesn’t set itself,” Stein said.
This hyper-self can also be a huge advantage for the company’s ad business. Stein announced a few weeks ago that Google “started some tests on ads within AI mode and within the Google AI experience.” Advanced targeted ads are the name of the game, seeing as Google’s biggest digital ad, Meta, announced it’s biggest hit recently.
Google’s business decisions will be relevant to the entire Web as the company’s products hold the potential to change the way we interact with the Internet. Their search engines are the user’s gateway to the Internet; Meanwhile, the company’s AI business is making major improvements along with competitors.
Personalization sounds good on paper, and promises to make life easier for many fans. But, like most technological advancements in our day and age, it comes with risks. The AI application knows more about you, the security risk of any data breach or transaction is there.
Even OpenAI CEO Sam Altman has acknowledged the dangers of allowing AI models to learn everything about you.
“There are two things that people are really interested in right now that taken together are a real security challenge. Number one, people like how these kinds of things get,” said Stanford University Professor Dan Boneh last month. “And then number two you can connect these types of other models.”
You can’t trust AI with this information any more than you can trust a human, Altiman said.
“If you tell your spouse a lot of secrets, you can trust that they will know from the manufacturer … what they will tell other people. Models don’t do this well. “So if you tell a model Yonke about your personal health problems, and then have them buy something from you, you don’t want an e-commerce site to know about all your health issues.”
